Shen, Lu; Mickley, Loretta, 2016, "Replication Data for: Influence of large-scale climate patterns on summertime U.S. ozone: A seasonal predictive model for air quality management”.

For inquiries about code and data, please contact Lu Shen (lshen@fas.harvard.edu)

To replicate all the figures in the paper, you will need to proceed as follows:

Step 1. Download all these datasets (All datasets are publicly accessible).
(1) NCEP Reanalysis (http://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html)
(2) 20 Century Reanalysis (http://www.esrl.noaa.gov/psd/data/20thC_Rean/)
(3) AMIP simulations 
(4) ERSST V3b (http://www.esrl.noaa.gov/psd/data/gridded/data.noaa.ersst.v3.html)
    HadISST (http://www.metoffice.gov.uk/hadobs/hadisst/)
    AMIP SSTs(http://www-pcmdi.llnl.gov/projects/amip/AMIP2EXPDSN/BCS/amipbc_dwnld.php)
(5) AMIP simulations (http://cmip-pcmdi.llnl.gov/cmip5/)
(6) AQS ozone (http://aqsdr1.epa.gov/aqsweb/aqstmp/airdata/download_files.html)

Step 2.Execute the R scripts (AMIP_East-T_SST.R) under ./AMIP/ (only necessary for Fig 5)

Step 3. Go to ./Figures and run the R scripts there.

The data structure is listed below.
./
Data
	20Century/(air.sfc.mon.mean.nc, prmsl.mon.mean.nc, slp.mon.mean.nc)
	AMIP/
	    (temperature and sea level pressure in 28 AMIP models)
	    1_AMIP_East-T_SST.R
	pressure(NCEP Reanalysis, at different pressure level)
	surface (NCEP Reanalysis, at the surface)
	SST(ERSST V3b, HadISST, and AMIP SSTs)	
Function/
	cov4gappy.R
	eof.mca.R
	filled.contour3.R
	get_geo.R
	get_met.R
	Henderson.R
	read_detrend.R
Figures/
	Figure 1/Figure 1.R
	Figure 2
		Figure 2.R
		Henderson_predict_O3.R
		mov_detrend_preict_o3.R
		Henderson.Rdata
		mov_detrend.Rdata
	Figure 3/Figure 3.R
	Figure 4/Figure 4.R
	Figure 5/Figure 5.R
